Bank of America CEO Brian Moynihan joined other Wall Street executives expressing concerns about the security risks posed by artificial intelligence (AI) models, specifically Anthropic’s Mythos system, Bloomberg reported on Thursday (July 16).
The evolution of AI technology has prompted the financial services industry and the United States government to assess potential cybersecurity threats to global markets, the report noted. Anthropic has said the Mythos model, unveiled earlier this year, excels at discovering vulnerabilities within digital infrastructures. Currently, Bank of America is among a group of Wall Street firms granted access to the platform, utilizing the AI model to test its own internal networks and share security information with industry peers.
The model has not been released to the general public. Earlier this week, JPMorgan Chase & Co. CEO Jamie Dimonwarned that granting broad access to the system would be equivalent to “giving ballistic missiles to individuals.”
In an interview broadcast Thursday on Bloomberg Television, Moynihan detailed the operational requirements these AI tools impose on financial institutions.
“It is a big change in the amount of work that will have to go on and the pace [at] which these tools will affect vulnerabilities in your systems, and how fast you have to take care of that,” Moynihan stated. “We are all working hard on it.” He further noted that the bank is monitoring external partnerships, adding, “We are making sure the third parties are doing the work and working cooperatively with the team.”
These warnings follow recent regulatory interventions regarding Anthropic’s technology. According to a July 1 report by PYMNTS, the U.S. Department of Commerce had previously enacted export controls and disabled access to Anthropic’s Mythos 5 and Fable 5 AI models. The export control directive cited unspecified “national security authorities” and concerns over a potential method for bypassing, or “jailbreaking,” the Fable 5 model. Notably, this specific security concern involved a vulnerability within Fable 5’s internal safeguards, which is distinctly separate from the Wall Street concerns regarding the Mythos model’s capability to uncover vulnerabilities in external systems.
The federal government recently lifted these restrictions after Anthropic worked with agencies to address the cybersecurity concerns. Following this regulatory approval, Anthropic announced it had restored access to Mythos 5 for a set of U.S. organizations. The artificial intelligence company noted it will continue working with the government to offer secure access to domestic and foreign partners involved in its Glasswing cybersecurity initiative.
The post Bank of America CEO Brian Moynihan Joins Wall Street Leaders in Warning of Mythos AI Risks appeared first on PYMNTS.com.
Source link
